The long-term goal of this project is to understand how m5C-cytosine methylases (m5C- methylases) achieve their exquisite specificity in recognizing DNA sequences. To achieve this goal Dr. Roberts is studying a set of enzymes that are related both structurally and functionally. Sequences are available for more than 50 m5C-methylases and recently a structure has been deduced for M.Hhal (recognition sequence: GCGC) both as a binary complex with its cofactor, S-adenosylmethionine, and as a ternary complex with DNA and the end product of the reaction, S- adenosylhomocysteine. From this structure a novel mechanism of DNA binding has been discovered that involves flipping the target cytosine residue completely out of the helix. Future experiments will use site- specific mutagenesis to probe the mechanism of this flipping in great detail, and to explore other aspects of the enzymology of M.Hhal. It is known that DNA recognition is mediated by a variable domain present in each of the m5C-methylases. A series of highly specific domain-swap experiments will be carried out with the idea of using MHhal as a crystallographic framework in which structural information may be obtained about other variable regions to discover if the mechanisms of DNA recognition are similar among this family of enzymes. Additional experiments will test whether base flipping is conserved among other DNA methyltransferases. Finally, protein engineering experiments will be undertaken aimed at modifying M.Hhal so that it can transfer groups other than methyl from the sulphonium center of analogs of AdoMet. These studies will provide information about DNA protein interactions and may help to widen the scope of usefulness of methyltransferases.